PhD Chapter 3

Results 3/3

This series of files compile all analyses done during Chapter 3:

All analyses have been done with R 4.0.3.

Click on the table of contents in the left margin to assess a specific analysis.
Click on a figure to zoom it

⏪ | 🏠

Sources of activity considered for the analyses:

Fisheries data considered for the analyses (expressed as number of fishing events or kilograms of collected individuals for each gear):

Gear Code Years Events Species
Dredge FishDred 2010-2014 21 Mactromeris polynyma
Net FishNet 2010 5 Clupea harengus, Gadus morhua
Trap FishTrap 2010-2015 1061 Buccinum sp., Cancer irroratus, Chionoecetes opilio, Homarus americanus
Bottom-trawl FishTraw 2013-2014 2 Pandalus borealis

1. Methodology

The aim of this section is to predict the structure of benthic communities based on the values of environmental variables.

We used abiotic parameters and indices of human exposure indices (calculated in Section 1) as predictors. We tested different methods: GLMs, GAMs, Random Forest and HMSC. Each method has been developed in dedicated scripts, whose final objects were imported here to present results and trends.

For each method, results are presented with a table regrouping McFadden’s or Tjur’s pseudo-R2, validation ratios and variables coefficients, and with maps displaying the probability of presence of each taxon for which the pseudo-R2 is higher than 0.20 (and different than 1). The raster presents results of the SDM (grey: low probability, dark blue: high). Stations are either plotted with colors (green = taxa present, red = taxon absent) or with circles (wider circles = higher taxon density).

2. Models

2.1. Generalized Linear Models

Diagnostics for each model can be found here and here.

2.1.1. Presence/absence data

We considered presence/absence data with a binomial distribution.

Abiotic parameters

Presence probability of significative taxa:

Prediction of specific richness based on this model:

Difference between predicted and observed taxa richness
station_id observed predicted difference
127 1 2 1
128 12 3 -9
129 15 3 -12
130 4 3 -1
131 10 5 -5
132 10 6 -4
134 9 7 -2
135 8 7 -1
136 9 8 -1
137 11 9 -2
138 12 9 -3
139 13 3 -10
140 14 5 -9
141 14 7 -7
142 22 6 -16
143 12 6 -6
144 12 6 -6
145 11 5 -6
146 10 6 -4
147 9 6 -3
148 20 7 -13
149 17 7 -10
150 11 5 -6
151 11 6 -5
152 15 4 -11
153 17 4 -13
154 11 4 -7
155 20 4 -16
156 22 3 -19
157 5 3 -2
158 19 4 -15
159 17 3 -14
160 14 4 -10
161 8 4 -4
162 11 3 -8
163 13 4 -9
164 14 4 -10
165 5 4 -1
166 7 5 -2
167 6 5 -1
168 19 5 -14
169 10 6 -4
170 22 6 -16
171 14 5 -9
172 10 10 0
173 18 10 -8
174 19 11 -8
175 13 11 -2
176 17 12 -5
177 9 11 2
178 19 10 -9
179 17 11 -6
180 12 12 0
181 21 11 -10
182 18 10 -8
183 6 10 4
184 12 11 -1
185 21 10 -11
186 14 9 -5
187 7 6 -1
188 5 45 40
189 16 3 -13
190 20 2 -18
191 9 3 -6
192 12 4 -8
193 2 4 2
194 7 18 11
195 9 3 -6
196 5 3 -2
197 13 3 -10
198 14 6 -8
199 7 3 -4
200 15 5 -10
201 12 7 -5
202 20 5 -15
203 19 5 -14
204 17 4 -13
205 12 7 -5
206 11 7 -4
207 17 6 -11
208 18 7 -11
209 5 2 -3
211 23 5 -18
212 4 7 3
214 24 6 -18
215 6 2 -4
216 6 3 -3
217 12 3 -9
218 17 3 -14
219 9 3 -6
220 11 3 -8
221 16 15 -1
222 17 13 -4
223 17 13 -4
224 11 11 0
225 21 11 -10
226 17 10 -7
228 6 4 -2
229 17 3 -14
230 17 4 -13
231 9 6 -3
235 19 8 -11
236 7 4 -3
237 14 6 -8
238 7 4 -3
239 14 4 -10
240 9 6 -3
241 12 3 -9
Exposure indices

Presence probability of significative taxa:

Prediction of specific richness based on this model:

Difference between predicted and observed taxa richness
station_id observed predicted difference
127 1 6 5
128 12 30 18
129 15 33 18
130 4 7 3
131 10 8 -2
132 10 30 20
134 9 28 19
135 8 29 21
136 9 17 8
137 11 16 5
138 12 26 14
139 13 8 -5
140 14 11 -3
141 14 19 5
142 22 14 -8
143 12 13 1
144 12 17 5
145 11 10 -1
146 10 13 3
147 9 17 8
148 20 42 22
149 17 43 26
150 11 43 32
151 11 43 32
152 15 43 28
153 17 43 26
154 11 43 32
155 20 43 23
156 22 43 21
157 5 8 3
158 19 10 -9
159 17 30 13
160 14 14 0
161 8 30 22
162 11 27 16
163 13 21 8
164 14 25 11
165 5 24 19
166 7 19 12
167 6 9 3
168 19 7 -12
169 10 20 10
170 22 7 -15
171 14 6 -8
172 10 17 7
173 18 26 8
174 19 19 0
175 13 20 7
176 17 27 10
177 9 22 13
178 19 20 1
179 17 27 10
180 12 21 9
181 21 29 8
182 18 27 9
183 6 23 17
184 12 25 13
185 21 26 5
186 14 27 13
187 7 28 21
188 5 5 0
189 16 7 -9
190 20 5 -15
191 9 42 33
192 12 31 19
193 2 12 10
194 7 28 21
195 9 4 -5
196 5 4 -1
197 13 7 -6
198 14 10 -4
199 7 43 36
200 15 13 -2
201 12 26 14
202 20 13 -7
203 19 13 -6
204 17 29 12
205 12 6 -6
206 11 17 6
207 17 9 -8
208 18 11 -7
209 5 4 -1
211 23 20 -3
212 4 7 3
214 24 10 -14
215 6 6 0
216 6 6 0
217 12 5 -7
218 17 8 -9
219 9 4 -5
220 11 4 -7
221 16 4 -12
222 17 5 -12
223 17 6 -11
224 11 8 -3
225 21 5 -16
226 17 5 -12
228 6 8 2
229 17 8 -9
230 17 6 -11
231 9 7 -2
235 19 42 23
236 7 41 34
237 14 8 -6
238 7 5 -2
239 14 44 30
240 9 43 34
241 12 44 32

2.1.2. Density data

⚠️ To be added … or not

2.2. Hierarchical Models of Species Communities

This section uses methodology and tools from Ovaskainen et al., with the direct help of Guillaume Blanchet.

First, we will compute models using the 108 stations with abiotic variables or exposure indices as predictors. 85 % of the stations (92) will act as training data, and the rest (16) will be used to validate the outputs. Second, these models will be used to predict taxa richness and distribution in the entire study area using predictor rasters.

We initiate the HMSC model with the chosen data:

  • presence/absence or density for dependant variable
  • exposure indices or abiotic variables for predictors

Priors and model parameters are set in the hmsc() function.

HMSC_PA <- hmsc(data, param, prior, family = "probit", niter = 100000, nburn = 1000, thin = 100)
HMSC_density <- hmsc(data, param, prior, family = "overPoisson", niter = 100000, nburn = 1000, thin = 100)

Here are the outcomes and diagnostics to evaluate each model’s quality (presented for each species seperately or averaged).

Diagnostics for each model can be found here and here.

2.2.1. Presence/absence data

We considered presence/absence data with a probit distribution.

Abiotic parameters

Mean of the predictor coefficients estimated by the HMSC model:

95 % confidence interval of the predictor coefficients estimated by the HMSC model:

Predictive power of the HMSC model:

Variance partitioning:

Presence probability of significative taxa:

Prediction of specific richness:

Difference between predicted and observed taxa richness
station_id observed predicted difference
127 1 2 1
128 12 2 -10
129 15 3 -12
130 4 4 0
131 10 5 -5
132 10 6 -4
134 9 9 0
135 8 9 1
136 9 10 1
137 11 10 -1
138 12 10 -2
139 13 4 -9
140 14 6 -8
141 14 9 -5
142 22 8 -14
143 12 9 -3
144 12 9 -3
145 11 10 -1
146 10 10 0
147 9 9 0
148 20 10 -10
149 17 9 -8
150 11 5 -6
151 11 8 -3
152 15 7 -8
153 17 3 -14
154 11 6 -5
155 20 3 -17
156 22 3 -19
157 5 2 -3
158 19 3 -16
159 17 3 -14
160 14 3 -11
161 8 2 -6
162 11 2 -9
163 13 3 -10
164 14 2 -12
165 5 2 -3
166 7 5 -2
167 6 3 -3
168 19 4 -15
169 10 6 -4
170 22 3 -19
171 14 3 -11
172 10 5 -5
173 18 6 -12
174 19 6 -13
175 13 9 -4
176 17 8 -9
177 9 9 0
178 19 10 -9
179 17 11 -6
180 12 11 -1
181 21 13 -8
182 18 11 -7
183 6 11 5
184 12 13 1
185 21 12 -9
186 14 11 -3
187 7 4 -3
188 5 8 3
189 16 3 -13
190 20 3 -17
191 9 2 -7
192 12 2 -10
193 2 5 3
194 7 7 0
195 9 2 -7
196 5 2 -3
197 13 2 -11
198 14 7 -7
199 7 8 1
200 15 7 -8
201 12 11 -1
202 20 3 -17
203 19 7 -12
204 17 4 -13
205 12 4 -8
206 11 5 -6
207 17 7 -10
208 18 9 -9
209 5 3 -2
211 23 5 -18
212 4 8 4
214 24 5 -19
215 6 8 2
216 6 1 -5
217 12 4 -8
218 17 3 -14
219 9 1 -8
220 11 2 -9
221 16 10 -6
222 17 9 -8
223 17 8 -9
224 11 8 -3
225 21 8 -13
226 17 8 -9
228 6 2 -4
229 17 3 -14
230 17 3 -14
231 9 4 -5
235 19 9 -10
236 7 7 0
237 14 4 -10
238 7 3 -4
239 14 8 -6
240 9 9 0
241 12 2 -10
Exposure indices

Mean of the predictor coefficients estimated by the HMSC model:

95 % confidence interval of the predictor coefficients estimated by the HMSC model:

Predictive power of the HMSC model:

Variance partitioning:

Presence probability of significative taxa:

Prediction of specific richness:

Difference between predicted and observed taxa richness
station_id observed predicted difference
127 1 2 1
128 12 7 -5
129 15 5 -10
130 4 3 -1
131 10 3 -7
132 10 5 -5
134 9 6 -3
135 8 5 -3
136 9 4 -5
137 11 5 -6
138 12 4 -8
139 13 3 -10
140 14 3 -11
141 14 3 -11
142 22 4 -18
143 12 4 -8
144 12 4 -8
145 11 4 -7
146 10 4 -6
147 9 4 -5
148 20 6 -14
149 17 7 -10
150 11 7 -4
151 11 7 -4
152 15 7 -8
153 17 6 -11
154 11 6 -5
155 20 6 -14
156 22 6 -16
157 5 2 -3
158 19 2 -17
159 17 3 -14
160 14 3 -11
161 8 4 -4
162 11 3 -8
163 13 6 -7
164 14 3 -11
165 5 3 -2
166 7 5 -2
167 6 3 -3
168 19 4 -15
169 10 4 -6
170 22 3 -19
171 14 3 -11
172 10 8 -2
173 18 9 -9
174 19 8 -11
175 13 5 -8
176 17 9 -8
177 9 4 -5
178 19 3 -16
179 17 8 -9
180 12 4 -8
181 21 9 -12
182 18 7 -11
183 6 4 -2
184 12 8 -4
185 21 6 -15
186 14 5 -9
187 7 2 -5
188 5 4 -1
189 16 3 -13
190 20 3 -17
191 9 6 -3
192 12 4 -8
193 2 5 3
194 7 5 -2
195 9 2 -7
196 5 4 -1
197 13 2 -11
198 14 3 -11
199 7 10 3
200 15 6 -9
201 12 6 -6
202 20 2 -18
203 19 3 -16
204 17 2 -15
205 12 3 -9
206 11 3 -8
207 17 3 -14
208 18 3 -15
209 5 2 -3
211 23 6 -17
212 4 5 1
214 24 8 -16
215 6 2 -4
216 6 2 -4
217 12 2 -10
218 17 3 -14
219 9 2 -7
220 11 2 -9
221 16 3 -13
222 17 5 -12
223 17 5 -12
224 11 4 -7
225 21 5 -16
226 17 5 -12
228 6 3 -3
229 17 2 -15
230 17 4 -13
231 9 4 -5
235 19 3 -16
236 7 5 -2
237 14 3 -11
238 7 4 -3
239 14 7 -7
240 9 5 -4
241 12 6 -6

2.2.2. Density data

⚠️ To be added … or not

2.3. Generalized Additive Models

2.3.1. Presence/absence data

⚠️ To be added … or not

2.3.2. Density data

⚠️ To be added … or not

2.4. Random Forest Algorithms

2.4.1. Presence/absence data

⚠️ To be added … or not

2.4.2. Density data

⚠️ To be added … or not

🔝